home / mcp / confluence mcp server

Confluence MCP Server

Provides Confluence REST API interactions: search, fetch, create, update, and delete Confluence pages via MCP.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "masna-ai-mcp-confluence-server": {
      "command": "mcp",
      "args": [
        "dev",
        "confluence.py"
      ],
      "env": {
        "CONFLUENCE_API_BASE": "http://localhost:8090/rest/api",
        "CONFLUENCE_PASSWORD": "your_password",
        "CONFLUENCE_USERNAME": "your_username"
      }
    }
  }
}

This MCP server lets you interact with Confluence Data Center through REST APIs by exposing a set of endpoints that AI models can call to query, create, update, and delete Confluence pages. It’s useful for automating content management, retrieving page data, and building workflows that integrate Confluence with other tools.

How to use

You run the MCP server through the MCP command-line tool in development mode. Start it and connect your MCP client to test and use the Confluence endpoints. The server exposes operations to search, fetch, create, update, and delete Confluence pages, so you can build natural-language prompts that translate into concrete Confluence actions.

How to install

Prerequisites: Python must be installed on your system. You also need the Python package manager to install dependencies.

Step 1: Install dependencies

pip install -r requirements.txt

Additional sections

Configuration is done by creating a .env file in the project root with your Confluence REST API details and credentials.

CONFLUENCE_API_BASE=http://localhost:8090/rest/api
CONFLUENCE_USERNAME=your_username
CONFLUENCE_PASSWORD=your_password

Running the server in development mode

Use the MCP CLI to run the server in development mode. This starts the MCP Inspector UI for testing and debugging.

mcp dev confluence.py

Direct execution alternatives

If you prefer to run the script directly, you can start the server with Python for basic testing, though this is not the standard deployment path.

python confluence.py

Installing in Claude Desktop

If you are using Claude Desktop, you can install this server for use within that environment.

mcp install confluence.py

Available tools

execute_cql_search

Execute a CQL query against Confluence to find pages that match the given criteria.

get_page_content

Retrieve the content of a Confluence page by its ID.

get_page_with_body

Fetch a page along with its body content for full page details.

find_pages_by_space

Find pages within a specific Confluence space by key, with optional expansion of properties.

find_page_by_title

Find a page by its title within a specific space.

create_page

Create a new Confluence page with a title, content, and optional parent.

update_page

Update an existing Confluence page with new content and optionally change its title or space.

delete_page

Delete a Confluence page by its ID.